This notebook analyses the TED talk data-set from Kaggle. The data-set can be obtained here. This notebook focuses on text-mining and demonstrates feature-extraction (using Term Frequency - Inverse Document Frequency), Topic Modeling (using Latent Dirichlet Allocation) and Named Entity Recognition.
Useful insights are gathered including favorite words used by speakers, different topics on which speeches are based, topic trends over time and most used persons in speeches.